Pathologic T stage and tumor volume predicted by PCA3 mRNA levels in urine of prostate cancer patients before radical prostatectomy

Abstract

LB-216

Background and Objective: Clinicians have been searching for more accurate prognostic information to guide their patients in choosing treatment for prostate cancer. PCA3 is a prostate specific non-protein coding gene that is over expressed in a majority of prostate cancers. Recent studies have reported diagnostic, as well as prognostic potential of PCA3 mRNA levels in post-digital rectal examination (DRE) urine. This study was conducted to assess the relationship between PCA3 mRNA levels in the post-DRE urine and subsequent clinico-pathologic features of tumors in whole mount specimens.
Patients and Methods: Post-DRE urine specimens were obtained from 65 men with prostate cancer before undergoing radical prostatectomy as their primary treatment. PCA3 and PSA mRNA levels were then measured in the whole urine using assays developed by Gen-Probe, Inc, San Diego, CA. PCA3 levels were normalized to PSA mRNA in each sample and the ratio was recorded as a PCA3 score. This score was then correlated with the comprehensive pathologic data from whole-mounted prostate specimens analyzed at the Armed Forces Institute of Pathology and compared with various other clinico-pathologic features.
Results: Patients with pT3 disease had much higher PCA3 score (median = 48.8, range =7.5-269.0) than pT2 patients (median = 18.7, range = 4.0-79.9), p = 0.02. Spearman analysis showed that PCA3 score significantly correlated with total tumor volume (n = 72, R = 0.38, P < .01). Patients with larger tumor volume (>2.0 cc) have significantly (p = 0.01) higher PCA3 score (median = 47.6, range = 7.5 - 269.0) compared to patients with smaller tumor volume (0.5 - 2.0 cc: median=17.5; <0.5 cc: median=18.7). Adjusted by other pre-surgical variables (age, race, serum PSA, clinical stage, biopsy grade, etc.), stepwise multivariate logistic regression analyses showed that PCA3 score was an independent predictor of pT3 prostate cancer (p = 0.01) and for a total tumor volume of <0.5 cc (p = 0.04). This is the first report of a urine assay predicting pathologic stage in prostate cancer.
Conclusions: This study shows that measurement of PCA3 score in the post-DRE urine of prostate cancer patients has potential to provide valuable prognostic information. Future studies addressing prognostic value of this assay are underway.